Abstract
A methodology for the prediction of low end and high end extremes in sustained water level is established. The observational data base is a long-duration sequence of monthly high and low extremes. The data identifies a deterministic trend attributable to Mean Sea Level rise and the nominal 19-year forcing in the astronomical tide. The data is pre-conditioned to remove these trends, defining a net data series suitable for extreme value analysis. Context-specific issues in the extreme value analysis are identified and resolved. These include probability model compatibility with elevation datums, rational estimation of the distribution parameters, and the estimation of confidence limits. The predictions for extreme low and high water levels are both real time and return period dependent.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
